Page 465 - Dust Explosions in the Process Industries

432 Dust Explosions in the Process Industries

a series of nomographs from which vent areas can be estimated,using Kst,the enclosure
volume, the maximum explosionpressure Pred, and the static openingpressure of the vent
cover P,,,, as parameters. The relevance of the “nomograph” method is tied to the nature
of the large-scale experiments on which it rests. These experiments were conducted
with dust clouds generated by blowing the dust into the experimental enclosures from
pressurized reservoirs through narrow nozzles, ensuring uniform, well-dispersed, and
highly turbulent dust clouds. Consequently, the burning rate for a given dust (see Chapter
4) was very high, in fact too high to be representative of the dust clouds in most indus-
trial situations.In spite of this, the nomograph method has been widely used. As discussed
in Section 9.3.7.5in Chapter 9, the conservativevent sizing approach of Verein deutscher
Ingenieure (1979) has been adopted as a basic also in a recent European CEN standard,
but venting area requirements may be eased if supported by adequate experimental
evidence.
Lunn, Brookes, and Nicole (1988) extended the VDI 3673 nomographs from 1979to
K,, values as low as 10bar m/s and Pred values down to 0.05 bar(g). The experiments on
which this extension was based were of the same kind as those forming the basis of the
original nomographs, that is, of very high turbulence and dust dispersion. Therefore,
the extended nomographs are subject to the same basic limitations as the original VDI
nomographs. This is illustrated in Figure 6.1, in which maximum pressurehent area cor-
relations predicted by the extended nomographs are shown together with experimental

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
VENT AREA Imzl

Figure 6.1 Experimental correlations between the vent area and maximum explosion pressure for
grain dust and starch explosions in a 2.8 m3 cubical vessel. Vents were covered with hinged metal
doors or glass panes in pivoted frames (From Brown and Hanson, 1933). Comparison with correla-
tions was suggested by Lunn (7 989).